Tom Lane wrote:
> Aside from the forward-reference problem, which seems easy enough to
> solve, I think there may be some performance issues that'd have to be
> dealt with (memory leaks and so forth).
OK, that's reasonable. If potential memory leaks are only for the
duration of a transaction I'm not worried (in regard to the one
recursive function we're using) as it's cachable and our trees not very
deep.
I have found one bug that crashed the backend if a recursive function's
called. Are you interested in it, since folks can now define them
semi-officially (if a bit unconventially) via CREATE AND REPLACE?
--
Don Baccus
Portland, OR
http://donb.photo.net, http://birdnotes.net, http://openacs.org